Although the nutrient requirements of all animals are, to a first approximation, the same, the feeding specializations of different species require that they obtain these requirements from foods varying widely in nutrient composition, caloric density, taste, packaging, abundance, and availability. Various morphological, physiological, and behavioral specializations have been evolved which permit animals to meet this problem of obtaining sufficient calories and a balanced diet. One behavioral mechanism is diet selection, in which the animal selects amounts and kinds of food times over relatively short intervals which, in conjunction with its physiological specializations, provide a calorically adequate and nutritionally balanced diet. Diet selection ranges from specialization in a single food to the consumption of a wide variety of foods in appropriate proportions. The means by which this occurs has been the subject of much experimentation, theorizing, and speculation in phychology, nutrition, and physiology and is gaining increasing attention in ecology. Most of the discussion on mechanism has focused on either taste or the feedback from the postingestive consequences. One set of variables which have been neglected are economic variables haveing to do with the animal's niche. optimality theory, developed by ecologists, has suggested that the maximization of caloric return from feeding is a basic principle. With a few exception, models based on optimality theory have concentrated on calories and ignored nutrient value. The present proposal is an attempt to develop a cost/benefit analysis of dietary choice using a laboratory simulation which has been successful in analyzing the problem of caloric return. If this analysis is successful, it should contribute to the discovery of the mechanisms of diet selection.